Risks of automated stock prediction engines

A celebrety tweeted about a stock tip to his 3 million followers. By the end of the day the stock price jumped over 290%.

"You can double your money right now. Just get what you can afford," Jackson tweeted about H&H; Imports, a money-losing venture out of Clearwater, Fla., that owns TV Goods, a marketing firm recently founded by Kevin Harrington.
While all the online reports I read assume that his followers fell for it, I am not sold on that. I would like to know if anyone ruled out the root cause as automated systems using twitter data for stock price change prediction. Automated systems using twitter data often use follower count to compute probability of some information being true. One of my news crawlers (scalebig.com) actually uses the same twitter feed to rate technical posts on scalability.


Knowing how widespread the use of twitter data is in different kind of automations, and because of how some of the way twitter data could behave is unpredictable, I would recommend looking at events like these with a microscope so that it may never happen again.

Comments

Popular posts from this blog

Chrome Frame - How to add command line parameters

Creating your first chrome app on a Chromebook

Brewers CAP Theorem on distributed systems